Linköping Studies in Science and Technology

Thesis No. 1793

Introducing New Materials in the Automotive IndustryManaging the Complexity of Introducing New Materials in Existing Production Systems

Fredrik Henriksson

Division of Machine DesignDepartment of Management and Engineering

Linköping University, SE-581 83, Linköping, Sweden

Linköping 2017

Copyright © Fredrik Henriksson, 2017

Introducing New Materials in the Automotive Industry - Managing the Complexity

of Introducing New Materials in Existing Production Systems

ISBN 978-91-7685-397-9ISSN 0280-7971

Distributed by:Division of Machine DesignDepartment of Management and EngineeringLinköping University

A true thing badly expressed becomes a lie

Stephen Fry

Du vet att verklighetenär inte alltid så verklig

Lars Winnerbäck – Dom Tomma Stegen

ABSTRACTPassenger vehicles are central to Western society, and contribute to a signifi cant part

of our greenhouse gas emissions. In order to reduce emissions, the automotive in-

dustry as a whole is working to reduce mass in passenger vehicles in order to reduce

energy consumption. One way to reduce mass is to introduce lightweight materials

in the body of the vehicle. This research aims to explore the relationship between

product and production system when introducing new materials.

Besides a theoretical review and an industry-centered technological mapping, four

case studies have been conducted during the course of this licentiate thesis. Two case

studies were conducted with engineering design students working as development

teams, one case study with the author as the developer and fi nally one case study in

an industrial environment at a product owning company with in-house production.

The goal of the case studies has been to increase the collective knowledge of how

product development decisions aff ect production development decisions, and vice

versa, when developing passenger vehicles in new materials.

In the following analysis of case study outcomes, a number of factors important

for introducing new materials are discussed. The relationship between product and

production is investigated, both in terms of how the production system aff ects the

product and how the product aff ects the production system. The outcome from this

analysis is a mapping of important factors for automotive industry companies to un-

derstand and identify when looking at introducing new materials in existing produc-

tion systems. Finally, a suggestion for future research eff orts is presented.

ACKNOWLEDGEMENTSLinköping, October 2017

The research presented in this thesis was performed at the Division of Machine De-

sign, Department of Management and Engineering (IEI) at Linköping University.

The research was performed within the Vinnova-funded Produktion 2030 strategic

innovation program.

I am grateful for all the support and help that I have received during the time of

writing this thesis, I really have been standing on the shoulders of giants in order

to see as far as possible. I would like to express this gratitude to several people, for

several diff erent reasons.

Firstly, I want to express my gratitude towards my main supervisor, Docent Kers-

tin Johansen, for discussions about what actually constitutes integrated product and

production development, how material, product and production are interconnected

and how to present my ideas in a way for others to understand what I am talking

about.

My former main supervisor and the person who employed me at Machine Design,

Professor Johan Ölvander, for giving me the chance to pursue my doctoral education,

giving advice, support and guidance as well as proofreading my eff orts during this

endeavor.

My second supervisor, Jonas Detterfelt, for active advise regarding research meth-

odology in engineering design, knowledge of the historical progress of the fi eld and

advice during the writing of this thesis.

I would also like to thank all interviewees, industry contacts and students who

have put up with my ideas and research eff orts. The data you contributed with when

performing your work was fundamental in this research, and without your ideas and

insights this thesis would not have been what it is right now. Thanks to all my col-

leagues, current and former, at the Division of Machine Design for supporting me

while researching and writing my thesis as well as creating an environment where

I’ve looked forward to coming into work (almost) every day.

Finally, I would like to thank my family: My parents for putting up with my ram-

blings about research projects, deadlines, interview questions and all other things

that have fi lled my head the last years. Thanks to my sister Frida, for helping me

make this thesis look at least semi-decent. Your help and support has meant a lot.

Fredrik Henriksson

APPENDED PUBLICATIONSThe appended papers are the basis of this thesis, and referred to as papers A – C in

the text. The papers are printed as they were originally published except for format

changes.

Paper A

Henriksson, F. & Johansen, K., 2016. An Outlook on Multi Material Body Solutions

in the Automotive Industry - Possibilities and Manufacturing Challenges. In: SAE

2016 World Congress and Exhibition. Detroit, United States, 12 - 14 April 2016.

Paper B

Henriksson, F. & Johansen, K., 2016. On Material Substitution in Automotive BI-

Ws-From Steel to Aluminum Body Sides. In: 26th CIRP Design Conference, Stock-

holm, Sweden, 15 - 17 June 2016.

Paper C

Henriksson, F. & Johansen, K., 2016. Including Student Case Projects in Integrat-

ed Product and Production Development Research – Methodology Description and

Discussion. In: 7th Swedish Production Symposium, Lund, Sweden, 25 – 27 October

2016.

ADDITIONAL PUBLICATIONSFollowing additional publications were published during the course of the research,

but are not included or discussed in this licentiate thesis.

Henriksson, F. & Johansen, K., 2014. Towards Applying The Boothroyd, Dewhurst

and Knight Methodology for Cost Estimation on Fibre Composite Manufacturing - A

Theoretical Approach. In: Proceedings of The 6th International Swedish Production

Symposium 2014, Gothenburg, Sweden, 16-18 September 2014.

Henriksson, F. & Johansen, K., 2015. Product development in the Swedish Automo-

tive industry: Can design tools be viewed as decision support systems? In: The 23rd

International Conference on Production Research, Manila, Philippines, 2 - 5 August

2015.

Henriksson, F.; Johansen, K.; Wever, R. & Berry, P., 2016. Student-developed labo-

ratory exercises - An approach to cross-disciplinary peer education. In: NordDesign

2016 - Highlighting the Nordic approach, Trondheim, Norway, 10 – 12 August 2016.

Kurdve, M.; Henriksson, F.; Wiktorsson, M.; Denzler, P.; Zachrisson, M. & Bjelke-

myr, M., 2017. Production System And Material Effi ciency Challenges For Large

Scale Introduction Of Complex Materials. In: Advanced Materials Proceedings, Vol-

ume 2 Issue 8 (492 – 499), 2017.

ABBREVIATIONSAbbreviation Meaning Unit

CFRP Carbon Fiber Reinforced Polymer -

ABS Acrylonitrile Butadiene Styrene (polymer) -

DFMA Design for Manufacturing and Assembly -

PDP Product development process -

FC Vehicle´s fuel consumption L/km

be Engine´s specifi c fuel consumption L/kWh

t Time s

v Instantaneous vehicle speed relative to ground m/s

Ft Tractive eff ort kN

η Drivetrain effi ciency N

FROLL Rolling resistance N

FACC Acceleration resistance N

FDRAG Aerodynamic drag N

FCLIMB Climbing resistance N

f Rolling resistance coeffi cient -

m Vehicle payload mass kg

g Gravitational acceleration m/s2

a Vehicle acceleration m/s2

CD Drag coeffi cient -

ρAIR Air density kg/m3

A Vehicle frontal area m2

v Instantaneous vehicle speed relative to air m/s

α Slope angle from horizontal degrees

CONTENTSPart 1 – Introducing the work

1.1 Introduction 1.2 Background 1.3 Aims and objectives 1.4 Research questions

Part 2 – Frame of Reference

2.1 Theoretical overview 2.2 Research methodology

Part 3 – Technology Mapping

3.1 Material Properties 3.2 State of the Industry 3.3 A general model for passenger vehicle production

Part 4 – Contribution

4.1 Industrial problem analysis 4.2 Case studies 4.3 Analysis and discussion 4.4 Conclusions 4.5 Future research

References

1

3 3 11 12

15

17 30

39

41 48 59

63

65 70 85 101 104

107

1

PART 1 - INTRODUCING THE WORK

2

PART 1 - INTRODUCING THE WORK

In Introducing the work, the introduction to and background for the work performed

are presented. This part should be read as an introduction to the research performed

and presented in the thesis, explaining the academic and industrial need for research

within the area of integrated product and production development as well as provid-

ing an understanding of the research questions posted.

Picture on previous page courtesy of author.

3

1.1 INTRODUCTION

“Why can’t we just switch the material?”

The question might seem valid, and we have probably all said it to ourselves at

one point in time during some project. Moreover, while it might seem easy to change

material in a product, there is a certain risk that the product would come out signifi -

cantly worse than before — if it were even possible to manufacture the product, that

is. Switching the material of a product or component is much more than changing a

drop-down menu in a CAD software application; instead, it necessitates further eval-

uation of both product and production properties to become a viable engineering de-

sign decision. Is it possible to create the selected geometry in the new material? Is it

economically viable? These are just some of the questions that need to be answered.

1.2 BACKGROUND

In 2015, the United Nations signed the Global Goals for Sustainable Development

[1]. These goals focus, among other things, on economic growth and environmental

sustainability [2], and a possible way of achieving these goals is to introduce more

sustainable materials in products that are distributed to the public in order to reduce

4

environmental impact while enabling economic growth.

Looking at the possible challenge of introducing new materials into products pro-

duced in existing production infrastructure, a signifi cant amount of research has

been done in the underlying research fi elds that form the basis of this fi eld; prod-

uct development research, materials and structural engineering and production en-

gineering research. This can be seen in diff erent applications, as academic and in-

dustry-based research has covered the design of specifi c components in lightweight

materials [3], material selection methodologies [4, 5], mass optimization [6] and

production development for new materials [7], but there is a research gap when

looking at the introduction of new materials in an existing production system. The

situation defi ned by introducing new materials in products produced in existing pro-

duction infrastructure is general, and can be seen in the automotive industry as well

as aerospace industry, consumer product industry and others. Looking at the impact

of mass reduction and multi-material design in the automotive industry, the research

was applied to this specifi c industry sector.

Automotive industry challenges

Personal transport, and the transport sector of the industry, are vital components

in Western society, where Sweden could be a suitable example. The transport sector

in Sweden contributes to 33% of the total greenhouse gas emissions (measured in

C02 equivalents) in the country, with road transports contributing to 93% of these

emissions; out of these 93%, passenger cars contribute to 64% of the emissions [8].

This means that passenger cars contribute to approximately 19.6% of the total green-

house gas emissions in Sweden. These numbers will diff er from country to country,

but it becomes evident that personal transport and passenger cars have a signifi cant

impact on emissions.

With a predicted growing middle class globally [9, 10], and thus an increased

5

availability of personal transportation (sources indicate a slight shift from work-re-

lated transport to leisure-related transport in Sweden [11, 12]), along with tightened

regulations on pollution and energy consumption, the automotive industry needs to

increase energy effi ciency in the vehicles it produces. One way of achieving this is to

reduce weight, or in some cases to keep weight at the present level while introducing

new functionality in the vehicle (to keep up with market demands). This could be

seen as increasing the “functionality over mass quota” for the product, or creating a

higher mass effi ciency in the product.

Increasing mass effi ciency could be solved by a more diverse choice of materials

for components and products, and this is a current trend in the automotive industry

with the multi-material design of many components (see the “State of The Indus-

try” section for further information). This increased diversity can be seen in fi gure 1,

showing material decomposition in American automobiles over the latest 40 years

[13, 14]. Notably, the plastic and plastic composites as well as aluminum content

have grown with time, while regular steel proposes a smaller portion of the materials

with each year (from ~54 per cent to ~35 per cent) [13, 14]. Looking at this shift for

the industry, it seems like advances in material and manufacturing technology, along

with increased demands on sustainability, enable new materials to be introduced and

used increasingly in mass-produced vehicles.

This material diversifi cation poses questions regarding the development of new

products: How can new materials be introduced more eff ectively, and how is the

material distribution managed when developing a new vehicle? If it is possible to re-

duce the implementation time and eff ort of introducing new materials the industrial

and environmental gain could be signifi cant, since more eff ective products could be

launched to market earlier.

To summarize, the automotive industry needs to introduce new materials to

6

reduce mass in their products, without sacrifi cing functionality or increasing the cost

outside the customer’s prize range. This also needs to be done continuously, and with

short lead time in order to be able to follow market trends as well as legal require-

ments.

Figure 1. Material distribution (in %) of an average American car by year (adapted from [13] and [14]).

0%

20%

40%

60%

80%

100%

Other materials

Lead

Textiles

Coatings

Other metals

Magnesium

Fluids and lubricants

Powder metal parts

Zinc die castings

Copper

Glass

Plastics/composites

Rubber

Aluminum

Iron

Other steels

Stainless steel

High-strength steel

Conventional steel

1977 1987 1995 2004 2010 2014

7

Production system challenges

To be able to introduce new materials, the production system needs to be taken

into account. The production system can be defi ned as “the people, equipment, and

procedures that are organized for the combination of materials and processes that

comprise a company’s manufacturing operations” [15]. The development of such a

system is called production development. Automotive production systems are com-

plex, complicated systems of systems that many times manage many diff erent prod-

ucts simultaneously, and do not necessarily have the same life cycle (neither in time

span nor place in time) as the products managed within the production system. In

dialogue with engineers working in the Swedish automotive industry, it has been

brought forward that the development of a production system is often more incre-

mental than that of a product, due to the size and complexity of the system; green-

fi eld production development (that could be compared to new product development)

in this fi eld is rare and very resource demanding. This means that the development

time for a production system is longer than that of a product, since design decisions

from earlier production system development are carried over in the physical confi g-

uration of the production system.

To summarize, the challenge for the production system is to enable both existing

and new materials, in future and present products, at an acceptable cost and without

the need for costly downtime while reconfi guring or investing in the infrastructure

(buildings, electricity, water, waste treatment, etc.) or production equipment.

Product development challenges

Before a product is introduced into production, a signifi cant amount of work is put

into developing it. Product development is, as defi ned by Krishnan and Ulrich, “the

transformation of a market opportunity and a set of assumptions about product tech-

nology into a product ready for sale” [16]. Product development processes are

8

therefore the sequencing of methods that companies use to bring new products

to a demanding market [17] where the goal is to “create a ‘recipe’ for producing a

product” [18].

A product development project can be described as a way of going from a scope

(market opportunity and assumptions about product technology in Krishnan’s and

Ulrich’s defi nition) through using data, tools and knowledge in a process to creat-

ing a result (as described in fi gure 2). Figure 2 explains how the factors scope, data,

knowledge and result in some way create boundaries for the process. The process

here would be the sequencing of tasks and decisions in order to move towards the

best possible solution (the result) given the scope, tools, knowledge and data at hand.

Figure 2. A generic model of a product development project.

Changing the material in the product would typically correspond to a change of

scope and result, as well as lower levels of data and knowledge regarding these new

product technology assumptions. Changing the material in the product while keep-

ing the same (or a very similar) production system would correspond to a partially

changed scope, a changed result, and lower levels of data and knowledge regarding

these new product technology assumptions due to path dependency as well as lost

or forgotten knowledge about decisions regarding the kept production system seg-

ments.

SC

OP

E

TOOLS

DATA

PROCESS

KNOWLEDGE RE

SU

LT

9

To summarize, product development engineers need to develop new and interest-

ing products, creating value for customers while still keeping production cost (both

in terms of unit cost and investment needs) within company profi tability margins.

This needs to be done more effi ciently, since product life cycles are decreasing while

product functionality is increasing.

An Industrial example

As an example of an industrial context, Volvo Cars AB is a company that needs to

manage the aforementioned challenges of the introduction of new materials in the

automotive industry, along with a change in product development work and in pro-

duction development work at its Torslanda plant in Gothenburg, Sweden.

Figure 3. Historic picture from Volvo Cars Torslanda [picture courtesy of Volvo Cars AB].

10

Volvo Cars AB have been a vehicle manufacturer since 1927, and have since then

been producing cars in Gothenburg, Sweden. The current plant in Gothenburg, Vol-

vo Cars Torslanda, was opened in 1964 [19] and has been developed and expanded

since. fi gure 3 shows how the plant looked in the late 1960s, while fi gure 4 shows a

more recent example of production at Volvo Cars Torslanda.

Volvo Cars Torslanda uses mixed-model assembly lines, which means that multiple

models with diff erent assembly requirements are assembled on the same production

line and with units of diff erent models mixed in sequencing [20]. As of May 2017,

four models were assembled at Volvo Cars Torslanda: the S90 II, V90 II, XC90 II and

XC60 II, all built on the SPA platform [21, 22, 23]. When the fi rst model built on the

SPA platform, XC90 II [21], was introduced, models built on earlier platforms were

still made at Volvo Cars Torslanda. This meant that the mixed-model assembly lines

had to accommodate both diff erent models and platforms.

Figure 4. Production of the second generation XC90 at Volvo Cars Torslanda [picture courtesy of Volvo Cars AB].

11

1.3 AIMS AND OBJECTIVES

The overall aim of this thesis is to enhance the product realization process by increas-

ing knowledge of the relationship between product and production development in

a certain context: the introduction of new materials into existing production systems

within the automotive industry.

The objective of this thesis is to increase the knowledge of how product develop-

ment decisions aff ect the production development design space, and vice versa, when

developing in new materials (or materials where the organization has very limited

prior knowledge) for existing production systems in the automotive industry.

The value gained for a company from a specifi c product (P) can be described as

[24][24]

where Q(t) is the rate of unit sales; p(t) is the unit revenue; c(t) is the variable unit

cost; F(t) is the design-specifi c fi xed cost; S(t) is the system cost (including, among

other things, overhead), support functions, and so on; and D(t) is the cost from prod-

uct and process development [24]. The eff ect of the research presented in this thesis

could be a reduction in D(t) in order to reduce risks of introducing new materials in

existing production systems.

12

1.4 RESEARCH QUESTIONS

In order to aim the research process towards the research objective, three research

questions were formulated. These three research questions correspond to the studies

performed. The conclusions should fi t the topic of introduction of new materials in

the automotive industry, and highlight diff erent aspects of this complex challenge.

RESEARCH QUESTION 1

How does the production system affect the introduction of new materials in products?

RESEARCH QUESTION 2

How well suited are current product development processes to manage the introduc-tion of new materials?

RESEARCH QUESTION 3

How could product and production development processes be altered to ease the introduction of new materials?

Structure of the thesis

The thesis has been divided into four distinct sections, as seen in fi gure 5: Introduc-

ing the work, Frame of reference, Technological mapping and Contribution. In Intro-

ducing the work, the background for the thesis is presented along with aims, objec-

tives and research questions. Frame of reference explains the theoretical framework

and research methodology that the research in this thesis is based on. In the section

Technological mapping, the author highlights material characteristics needed to ex-

plain the technological challenge of introducing new materials in existing production

13

Figure 5. The main structure of the thesis.

12

34

INTRODUCING THE WORK

FRAME OF REFERENCE

TECHNOLOGICAL MAPPING

CONTRIBUTION

systems, explores the state of the industry and presents a generalized model of car

production. In the fi nal section, Contribution, a state-of-the-industry review is pre-

sented along with a theoretical analysis and descriptions of four case studies per-

formed. These six data sets are then analyzed to formulate conclusions to the re-

search questions presented in the fi rst section.

15

PART 2 - FRAME OFREFERENCE

16

PART 2 - FRAME OF REFERENCE

In the frame of reference, the theoretical framework is presented along with the

methodology used in this thesis. This part should be read as a benchmarking of ex-

isting work in the academic arena, especially regarding the research fi elds of product

development and production systems.

Picture on previous page courtesy of Philip Ekströmer.

17

2.1 THEORETICAL OVERVIEW

The theoretical overview will cover two domains: product development and produc-

tion systems. Some eff ort will also be put into exploring the overlap between the

domains that can be seen in fi gure 6.

Figure 6. A visualization of the theoretical overview.

PRODUCTDEVELOPMENT

PRODUCTIONSYSTEMS

18

Product development

Product development is, as defi ned by Krishnan and Ulrich, “the transformation of a

market opportunity and a set of assumptions about product technology into a prod-

uct ready for sale” [16]. Product development processes are therefore the sequencing

of methods that companies use to bring new products to a demanding market [17],

with the goal to “create a ‘recipe’ for producing a product” [18] that can increase the

company’s revenue. The impact of the product development process on the resulting

product is debated: some authors have proposed a clear connection between sys-

tematic product development processes and commercial gain [25, 26], while others

suggest that it is not the methods but rather the interactions between actors in the

organization that undertake the development project that creates profi table prod-

ucts, and that what has been contributed to the methodology by other authors has

been confused with the result of the process [27]. This has analogies in work on

software design, suggesting that the rational design process is at best a work of fi c-

tion, but that there might be reasons to fake a rational design process by creating

documentation suggesting such [28]. Others have reported on a connection between

the usage of formal methods and development time (shortened) and project failure

rate (reduced) [29].

With a cross-disciplinary topic like product development, diff erent perspectives

become apparent in academic publications. At least four common – and diff erent -

perspectives on product development exist: marketing, organizational, engineering

design and operations management [16]. These perspectives also aff ect which type of

knowledge regarding the product development process is presented in the research.

Looking specifi cally at the interface between design and manufacturing, Dekkers et

al. [30] present six reoccurring themes on the interface stemming from these per-

spectives on product development: a) integral productivity, b) order entry points and

modularity, c) product life cycle management, d) sourcing decisions

19

and supplier involvement, e) integrated processes and coordination, f ) enabling

through ICT” [30].

Product development processes

Product development processes can be divided into two types of process models:

linear or quasi-linear process models and non-linear process models. In linear or

quasi-linear process models there is a clear sequencing of tasks, and while iterations

might be mentioned in the description of the process, they are not emphasized nor

described in the visualizations of the process models. The iteration and lack of se-

quencing is more articulated in the non-linear process models, some of which explic-

itly describe a working process with no self-evident order of tasks and many itera-

tions before a product is fi nished.

Linear or quasi-linear process models

Linear or quasi-linear process models emphasize a sequential dependence between

diff erent activities within a product development project. Established variants of the

linear or quasi-linear process have been introduced by Ulrich and Eppinger [31],

Ullman [32], Cooper (with the stage-gate model) [25, 33, 34, 35] and the Systematic

Engineering Design community led by Pahl, Beitz, Hubka and Eder [26, 36]. Linear

product development process models are restricted by their lead times and the built-

in sequential approach to development tasks [17].

Ulrich and Eppinger [31] describe a simplifi ed, linear model of the development

of new products (as seen in fi gure 7), with focus on the types of thinking involved

(divergent or convergent). Focus is on the three functions in a company central to the

development of new products: marketing, design and manufacturing [31]. Ulrich

and Eppinger suggest six important and distinct phases in the development of new

products [31], something that is reinforced by Ullman [32]. Cooper suggests fi ve

stages instead of six [25]. The Systematic Engineering Design community

20

is split on how many phases a product development project should be divided into;

Hubka and Eder suggest six phases [36], while Pahl and Beitz propose fi ve [26]. Ap-

parent in all descriptions is that the product development process can be divided into

distinct and identifi able phases with diff erent activities, tools and objectives. Instead

of referring to diff erent phases, one can talk about development activities and what

happens within these activities.

An interesting anomaly is that the stage-gate model does not start with a user

problem of some sort, but instead begins with a product idea. This means that the

following feasibility analysis and customer/user research stage is colored by the ini-

tial product idea. Other linear product development process models start with a user

problem.

Figure 7. The product development process as described by Ulrich and Eppinger [31].

21

Common for all of the linear product development process models is that they do

not propose a plan of action for ambiguous input, especially if the input is some sort

of analysis based on output from the process itself. . The processes are designed to

progress, with the next phase or activity building on a clearly defi ned input which

comes from the output from an earlier phase or activity. This implies that work might

halt in a linear process if the result of an activity is “maybe” instead of a clear “yes” or

“no”.

Non-linear process models

Non-linear process models do not have a common view on the progression of the

product development process, but all stray away from the linear and sequential pro-

gression model described by, for example, Ulrich and Eppinger [31]. Concurrent

engineering, as coined by Prasad [37], is described as two sets of technological and

organizational factors that require integration and development: integrated prod-

uct and process organization and integrated product development. Concurrent en-

gineering principles have been used to integrate product and production systems

development in some cases [38], then often coupled with systems engineering prin-

ciples and their focus on holistic development [39], as seen in Figure 8. Concurrent

engineering could be used to develop products and production systems in collabora-

tion if both are modeled as a combined system. Systems engineering could be used

to describe such a complex system, although this requires access to detailed data on

a large number of parameters.

Another way of approaching non-linear development is to adapt Agile develop-

ment strategies. The Agile development methodology originates in software devel-

opment [40], and works around an iterative and incremental process where changes

in requirements are encouraged in order to maximize the relevancy of the fi nished

product [41]. One of the main goals is to have a functioning product at the end of

each iteration [41].

22

Figure 8. Systems Engineering as described by Prasad [37].

Analysis of product development process models

While all the evaluated product development process (PDP) methodologies address

production, there is no real consensus on how the production system should be de-

veloped at the same time as new products are developed. The methodologies men-

tioned in this thesis still view the production system as a fi xed system with a set of

restrictions that could be adhered to or discarded, but not reworked or abstracted as,

for example, customer demands.

The PDP methodologies described as linear (Ulrich & Eppinger, stage-gate mod-

els) seem to be more challenging to combine with an integrated model of product

and production development. This could be due to the iterative nature of an integrat-

ed product and production development process, but cannot be solely

SYSTEM ANALYSIS &CONTROL (BALANCE)

PROCESS OUTPUT

DESIGNSYNTHESIS

PROCESS INPUT

REQUIREMENTSANALYSIS

FUNCTIONAL ANALYSIS& ALLOCATION

REQUIREMENTSLOOP

DESIGNLOOP

VERIFICATION

23

contributed to this factor. While visual representations of the process are important,

they do not give the full representation of the authors’ line of thought on the product

development process.

Neither the linear nor non-linear models are perfectly suited for integrating prod-

uct and production system development, since the linear models have issues with

ambiguity in outputs from activities, while the non-linear models require either large

quantities of reliable data reducing the opportunity for innovation or require an ex-

tremely modular output from the process, more akin to software than integrated sys-

tems of software and hardware.

Product modularity, platforms and architecture

In order to rationalize the development and manufacturing of multiple products and

variants in the same manufacturing facility resulting in a higher number of possi-

ble product variants for the same cost and a reduced time to market [42], prod-

uct-producing companies have developed modularity strategies for products and

product families. Product modularity means that the product is divided into several

sub-groups or sub-systems [43], with interfaces between modules. Typical interfaces

are attachment interfaces, spatial interfaces, transfer interfaces, control/communi-

cation interfaces, user interfaces and environmental interfaces [44]. These modules

can then be integrated into products in diff erent manners; within a company, the

manner of integrating modules and components is called an architecture [45]. Ar-

chitectures can be both highly modular (with a high level of interchangeability be-

tween modules) and highly integrated (with a low level of interchangeability between

modules) [46]. According to Shibata et al. [47], as technology matures, product and

technology-related knowledge increases and market demands change, most product

segments evolve from an integrated architecture with complex interfaces towards a

more open modularity with less complexity in interfaces

24

and in the connection between structures and functions.

Since components can be divided into standard segment components, brand-iden-

tifying components and specifi c model components [43], product-producing com-

ponents use this to group components of the same class into modules in order to

account for future design updates [48].

Product platforms are combinations of an architecture and standard segment or

brand-identifying components and modules that are shared across a range or family

of products [49]. Platforms are used to reduce development time for new products

while adding capabilities to upgrade or introduce new products [50] and lower var-

iable cost for subsystems [49], but can generate a level of overdesign for low-end

products and are not suitable to implement if the market diversity is too low or too

high [49]. Some researchers have suggested an expansion of the product platform

topic to further include production processes [51].

Design for manufacturing and assembly

Design for manufacturing and assembly, or DFMA, is a structured approach to prod-

uct development in order to minimize waste (in time, cost and material) in produc-

tion and to create effi cient products. Ulrich et al. [24] describe the goal of DFM, a

part of DFMA, as to make a product easy to manufacture during the design phase of

the product development process. Two important objectives in DFMA are reducing

the number of components and thus reducing processing and assembly, and design-

ing the remaining components in order to simplify processing and assembly [52].

Value engineering, the practice of questioning material and process selections in or-

der to fi nd solutions with lower costs while still providing a comparable functional-

ity, is also an important part of DFMA [53]. One of the main changes of the design

process is to spend more time in the conceptual design phases in order to reduce time

needed to change design due to manufacturing restrictions [54].

25

DFMA is one way of mitigating the risks of over-the-wall design approaches, ,

where the design engineer fi nishes the design on their own and delivers it to the

manufacturing engineer to fi gure out how to produce [54]. The over-the-wall design

approach was the traditional way of developing new products when the complexity of

products had exceeded what one person could manage, but before structured ways of

performing tasks within the development process simultaneously or in parallel had

been developed [32].

Production Systems

All cars, and other products, have at some point in time been made into a product.

The terms manufacturing and production are sometimes used interchangeably for

the activity of converting raw materials into products. Kalpakjian and Schmid [55]

describe this as manufacturing, and defi ne the activity as “the process of converting

raw materials into product.” Merriam-Webster defi nes production as “the creation of

utility; especially: the making of goods available for use” [56], and manufacturing

as “the process of making wares by hand or by machinery especially when carried

on systematically with division of labor” [57], so these two words have very similar

meaning in the English language. Therefore, the author has chosen to defi ne produc-

tion and manufacturing within the work of this thesis.

Production is, in this thesis, defi ned as the umbrella term for all activities involved

in transforming the product from a digital or physical description of the product, its

components and how they are interfacing each other, into a physical entity ready to

be sold and used. In this thesis, production consists of manufacturing and assembly

(as seen in Figure 9), and manufacturing in itself consists of activities such as form-

ing, cutting, joining, casting and surface treatment, as illustrated in Figure 9. These

descriptions of manufacturing and assembly are very similar to what Boothroyd et.

al use in their base for DFMA [54].

26

While joining is present both as a manufacturing and assembly process, perma-

nent joining would generally be considered a manufacturing process and non-per-

manent joining an assembly process, but there are exceptions to this classifi cation.

This aforementioned defi nition of production does not take software changes into

account. This is based on the view that a software installation or change is a ration-

alization of a production process (diff erent hardware with implemented software for

each variant) into the reconfi guration of the software. As an example, the engine

control unit in a line of cars could either be dedicated and have diff erent hardware

specifi cations for each power output, or have generalized hardware and use software

reconfi guration in order to diff erentiate the product. In this thesis, the fi rst scenario

is assumed to be the case within the production process.

Figure 9. The terminology of production used in this thesis.

All product-producing companies have some sort of production system, in-house.

The production system can be defi ned as “the people, equipment, and procedures

that are organized for the combination of materials and processes

PRODUCTION

MANUFACTURING ASSEMBLY

FORMING CUTTING JOINING

SURFACETREATMENT

CASTING &MOLDING

ADDITIVEMETHODS

27

that comprise a company’s manufacturing operations” [15], and production develop-

ment should be seen as the development of the system.

Manufacturing

In this thesis, manufacturing is used to describe the physical realization of compo-

nents from raw material to fi nished component for assembly. In metal manufactur-

ing, the manufacturing has been classifi ed into diff erent activities such as casting,

powder metallurgy, plastic forming, cutting, machining, unorthodox machining or

cutting, joining and surface treatment [58]. This would be similar for other materi-

als, with smaller changes (casting in polymers is referred to as molding, as an exam-

ple). For composite materials, casting is replaced with molding, and plastic forming

is unusual with the exception of sheet molding compound, SMC [59], that is in-

troduced in the manufacturing process as a sheet and then formed to manufacture

components.

The three most defi ning parameters of manufacturing processes are function,

quality and cost. These three parameters are in confl ict with each other, as the lowest

cost cannot be combined with the highest quality and highest functionality [60].

Assembly

Assembly is the activity consisting of joining components together into systems or

products [61]. This is usually the fi nal part of the production process, and therefore

dependent on upstream activities in the production process [61].

The assembly process can be organized in several diff erent ways, with job shop

assembly [62], single-model continuous assembly lines [62], batch assembly lines

[62] and mixed-model assembly lines [63] being some of the diff erent assembly

process organizations. The organization of the assembly process aff ects the level of

product variance in each work station and the sequencing of components assembled

to the product.

28

Production development

While products and production equipment are developed simultaneously in industry

today, the development processes are not integrated but very much independent, bar

some specifi c milestones where the projects are synchronized [64]. Development of

the production system can be divided into two categories, technology development

and production development, where technology development is development of the

production system applied to existing products and production development is de-

velopment of the production system applied to new products (as described in fi gure

10) [65].

Figure 10. The Product-Production mix as described by Schätz [65].

Bellgran [61] describes a development process for the assembly system (a subset

of the production system) that includes two main elements: preparatory design and

design specifi cation. The preparatory design consists of the early design activities

with low investment density, while the design specifi cation element is focused

PRODUCTDESIGN

PRODUCTIONDEVELOPMENT

CONTINUOUSIMPROVEMENT

TECHNOLOGYDEVELOPMENT

NE

WO

LD

OLD NEW PRODUCTION

PR

OD

UC

T

29

on complexity management, strategy and design criteria [61]. Schätz [65] describes

production development, both at the cell and system level, as a part of the product

development process. Traditionally, production development has been performed

after the product design is fi nished in a sequential, over-the-wall manner, but new-

er concurrent approaches have been applied to production development as well as

product development in general [37, 65], while others discuss the co-evolution of

products and production systems [66].

Flexibility in production

Production fl exibility is a measurement of how adaptive the production system is to

change [67], and is derived from the defi nition of fl exibility as the ability to manage

changing circumstances and uncertainty [68]. These changes can be due to internal

(such as small batch volumes) or external (such as fl uctuations in market demand)

factors [67, 69], but only within a predefi ned range of values [15]. A production sys-

tem can be fl exible in diff erent ways, called dimensions [70]. D’Souza and Williams

[70] defi ne the four fl exibility dimensions of volume, variety, process and materials

handling. These dimensions consists of several measures [70] including volume fl ex-

ibility [67, 69, 70], mix fl exibility [67, 69, 70] and changeover fl exibility [67, 69, 70].

Slack [71] instead proposes the dimensions of product, mix, volume and delivery

fl exibility.

All fl exibility dimensions should not be fully utilized at the same time in the same

production system [15, 72, 73], and according to Slack [71], managers in industry

often seek to limit the need for fl exibility since fl exibility can be viewed as a cost (i.e.,

underutilized equipment and facilities and waste). This requires a review of the tasks

at hand and a ranking of prioritized dimensions of fl exibility [69]. A review of the

level of uncertainty aff ecting the decision should also be performed in order to assess

the actual need of fl exibility [68].

30

2.2 RESEARCH METHODOLOGY

In this chapter, the research methodology of this thesis is presented. First, the gener-

al process is described to give context to the structure and reasoning behind the ap-

proach taken. Second, the two main themes of the research (case studies and action

research/participation action research) are further described.

General process

The general research process for the work presented in this thesis consists of litera-

ture studies, observations and interviews, along with four case studies. The four case

projects were performed sequentially with no overlap between the projects them-

selves, as shown in fi gure 11. The fi rst two case projects were performed on student

projects (shown in orange in fi gure 11), the third had the researcher as the design

engineer (shown in purple in fi gure 11) and the fourth had the researcher (again in

purple) being the project coordinator within an industry project (engineers in indus-

try shown in grey in fi gure 11).

Outcomes from the literature study formed the basis for the case projects, along

with observations and interviews in industry. Each case was fi rst briefl y analyzed

before setting up the next case study.

Documentation from each case has been compiled and used to conduct a com-

bined analysis along with the current state of the industry, as well as a theoretical

review to understand and describe how integrated product and production develop-

ment is performed, as seen in fi gure 12.

The multiple-case study, in combination with industrial observations and litera-

ture reviews, is a deviation from the main research methods on integrated product

and production development, that are survey studies and single-case studies [30].

While case studies are explicitly more descriptive and suitable for explaining a set of

decisions [74] as well as the social interactions that are the core aspects

31

Figure 11. Temporal disposition of the work presented in this thesis.

CA

SE

1C

AS

E 2

CA

SE

3

TE

CH

NO

LOG

ICA

L M

AP

PIN

G A

ND

TH

EO

RE

TIC

AL

RE

VIE

W

OB

SE

RV

AT

ION

S, I

NT

ER

VIE

WS

ETC

AN

ALY

SIS

CA

SE

4

32

of engineering [75], as Dekkers et al. [30] interpret Timpf [76], there is a risk in

making valid generalizations from a single-case study. Since many interactions or

occasions are unique within a single case, the basis for generalization is limited. This

could be mitigated by aggregating several cases into a multiple-case study. A series of

case studies based on industrial challenges with analysis and further development of

the next case in collaboration with industry is similar to the “industry as a laboratory”

research approach, initially presented by Potts [77] and further expanded by Björns-

son [7, 78]. Cases have been designed as student team case projects, researcher case

projects and industrial participation action research projects, further diff erentiating

the data input in order to generate generalizable knowledge.

Figure 12. Structural disposition of the work presented in this thesis.

Ottosson & Björk [79] have proposed an inverse relationship between the distance

between researcher and the object of research, and reliability of research

CASE STUDIES

CASE STUDIES ON STUDENT PROJECTS

CASE 1 - CARBON FIBER ROOF PANEL

CASE 2 - ALUMINUM ROOF PANEL

RESEARCHER CASE STUDY

CASE 3 - ALUMINUM BODY SIDE

INDUSTRIAL CASE STUDY

CASE 4 - ALUMINUM ROOF PANEL II

THEORETICALREVIEW

TECHNOLOGICALMAPPING

ANALYSIS

CO

NC

LUS

ION

S

33

(as depicted in fi gure 13). This implies that the researcher should have an active role

in the cases performed, rather than a passive and distant one. In an opposite take,

the research eff orts identifi ed by Ottoson & Björk [79] as having the largest distance

between the researcher and the object or occasion researched are the research eff orts

most replicable. In order to ensure a high level of research quality, by ensuring rep-

licability (and thus enabling falsifi cation) as well as reliability, a combined approach

using research activities with both high reliability (short distance between researcher

and the object of research) and high replicability (long distance between researcher

and the object of research) is a suitable way of creating a general research methodol-

ogy in order to generate generalizable and useful knowledge.

Figure 13. A fi gure of how the distance to the object or occasion researched negatively affects reliability, adapted from [79].

RE

LIA

BIL

ITY

DISTANCE FROM OBJECT

OWN USE / PARTICIPATION

SIMULATIONS

OBSERVATIONS

DIALOGS

STRUCTURED INTERVIEWS

34

Case study research

A case study is a form of descriptive, qualitative research, more aiming to answer

questions relating to “how?” and “why?” than for example “what?” or “how much?”

[74]. Case studies have been shown to be useful when the object or environment

under study is hard to defi ne or “messy” [80]. Combined with the suitability of case

studies to investigate complex events [81], case studies have been popular in both

product and production development research. Examples are implementations of

new development approaches [82], validations of design tools [83], investigations

of material effi ciency [84], generating possible concepts for novel human-robot col-

laboration systems [85], data gathering regarding assembly fl exibility [81], the eff ect

of responsiveness and fl exibility to customers on manufacturing [86] and comparing

strategies for competitiveness in manufacturing companies between countries [87].

A project in engineering design in general, and student projects in particular, can

be described as a way of going from a scope, via using data, tools and knowledge in

a process, into creating a result (as described in fi gure 14). In fi gure 14, there is an

interaction between scope, knowledge, tools, data and process in order to create the

best possible result with available resources. In the case of an engineering design

project, the process is the sequencing of activities to progress from scope to result.

If the process is to be studied, the process itself cannot be validated with industrial

examples. What could be done instead is to create the project scope so that it is in-

dustry-like and has industrial relevance, and so that the data the case group gets is

relevant and industry-like. It is also important to identify diff erences in knowledge

between students and engineers in industry, provide tools similar to those in industry

and evaluate whether the result is reasonable from an industrial point of view.

35

Figure 14. A generic description of an engineering design project.

Since the aim of the research presented in this thesis is not to exactly identify the

process that is used in industry, but instead to generate in-depth knowledge about

the types of challenges that can occur in these types of projects, a process aligning

reasonably with what is done in industry is needed in order to consider the data in-

teresting and signifi cant.

If the scope in fi gure 14 is developed in collaboration with industry, and the result

is evaluated by experts from industry, the input and output of the project can be regu-

lated and deemed as similar enough to industry projects. If an industrial contact can

also provide real or realistic data, the probability of a realistic result increases and the

process is furthermore boxed in. While there are diff erences in how students and in-

dustry professionals approach certain challenges, knowledge from a case project on

students can be transferred to be useful in industry if the project has an appropriate

scope, is set up properly (with regard to data, tools and knowledge) and the result

from the project can be deemed reasonable. If these three properties exist, the pro-

cess studied in the case study is similar enough to an industrial process to generate

valuable knowledge, even if the case participants are students.

SC

OP

ETOOLS

DATA

PROCESS

KNOWLEDGE RE

SU

LT

36

Multiple data collection methods should be used, preferably with mandatory

course assignments and other teaching activities as a part of the data collection to

reduce the risk of participants focusing more on delivering data than performing the

project in the case. Since there are diff erences in knowledge between engineering

students and engineers in industry, a multiple-case study combining cases is recom-

mended in order to maximize data to analyze.

Specifi c qualities in the case that aff ect data collection regarding the work process

are the project scope (how well this resonates with industry-based projects), knowl-

edge (of the participants in the project), data given as input to the project, tools used

in the project and the result of the project. The least robust qualities are the input

data and the project result.

Action research and participation action research

Action research, originating from the fi eld of educational research [89], is a meth-

odology where the researcher is participating in the subject that is being observed

[90]. Simplifi ed, action research can be described as a series of look-think-act [88]

(as seen in fi gure 15) or action-refl ection loops ( [91] via [89, 90]), as seen in fi gure

16.

Figure 15. The look-think-act loop of action research, adapted from [88] ..

LOOK

THINK

ACT

LOOK

THINK

ACT

LOOK

THINK

ACT

LOOP ILOOP II

LOOP n

37

Like case study research, action research is also considered a form of qualitative

research [88] but with the researcher more immersed in the object of study. The

research work focuses on process rather than results, and views the researcher as a

catalyst for the work process [92].

Figure 16. The action-refl ection loop as described by [91] and adapted from [89].

Participation action research (PAR) is an expansion of the action research method-

ology specifi cally intended to generate knowledge regarding development processes

[93]. In PAR, the researcher works actively as a manager managing a development

project of the kind that is to be researched; this is done to immerse the researcher

in the complexity of a development process and gather implicit knowledge as well as

unspoken information used by developers in the process [93]. Ottosson [93] propos-

es a participation action research approach, but highlights the risk of the researcher

becoming too immersed in the development environment and losing the scientifi c

approach to the development process.

A solution to this could be to have a scientifi c environment present during the pro-

ject where the researcher can retract and refl ect as the project progresses,

ACTION REFLECTION

38

as presented in fi gure 17. In this thesis, a combination of case studies using participa-

tion action research, literature and industry reviews and industrial observations has

been used to gather data. The case studies have employed diff erent levels of active

participation, with Case 3 having the highest level of participation (the author per-

forming the work), Case 4 the second-highest (the author acting as a coordinator),

and Cases 1 and 2 the lowest (the author acting as a supervisor).

Figure 17. The difference between participation action research (the lower ellipse on the right) and tradition-al observational research (the upper ellipse on the right), adapted from [93].

“TRADITIONAL OBSERVATIONAL RESEARCH”

PARTICIPATION ACTION RESEARCH

INTERNAL DIALOG AND DISCUSSION

39

PART 3 - TECHNOLOGI-CAL MAPPING

40

PART 3 - TECHNOLOGICAL MAPPING

In the technological mapping, material properties and industrial implementation of

multi-material design are presented, followed by a general description of vehicle pro-

duction. This part of the thesis should be read as a presentation of possible materials

to combine and introduce, with a brief explanation of diff erences in properties, as

well as an industry-review of multi-material products in order to understand the

state of the industry. Parts of the material presented in this section is also presented

in Paper A.

Picture on previous page courtesy of Lantmäteriet (Ekonomiska kartan 1956, kartblad Iggesund J133-15H7d59)

41

3.1 MATERIAL PROPERTIES

Material properties are descriptions of how materials behave under certain specifi c

conditions. These behaviors are fundamental in the selection of materials in mechan-

ical engineering, since this determines whether the product will fulfi ll the require-

ments set up for it. Common materials for automotive components are metals (either

in steel alloys or aluminum or other lightweight alloys) and polymers (sometimes in

the form of polymer composites).

Figure 18. Specifi c strength vs. specifi c stiffness for polymers, composites and metals [data courtesy of Granta Design Ltd].

42

Two interesting material properties are specifi c strength and specifi c stiff ness. In

combination, these two describe how a material can be manipulated without ruptur-

ing, something that is of great interest when evaluating materials for panel forming

and later use and risk of blunt force impact during use (impact from opening doors

at a parking lot, tools, etc.). Materials with both high specifi c strength and specifi c

stiff ness are more durable to smaller “dings” or bumps, since they will not deform

as much as weaker materials. In fi gure 18, polymers (the blue grouping) is shown to

have both lower specifi c strength and specifi c stiff ness than metals (red, green and

magenta groupings). This implies that the polymers are less likely to withstand small

bumps without leaving marks.

Two other parameters related to mechanical manipulation of the material are yield

strength and tensile strength. Yield strength is the stress the material can manage

without plastic deformation [94], and tensile strength is the stress the material can

manage before necking and following catastrophic failure [95].

Figure 19. Specifi c stiffness over specifi c strength for material classes. A higher specifi c strength but similar specifi c stiffness can be seen for composites (compared to metals and alloys). [96]

43

Materials with a tensile strength very close to yield strength are called brittle, since

they do not deform before failure [94]. In fi gure 20, the blue grouping showing pol-

ymers has both lower yield and tensile strength, and has a slightly higher Y/X coor-

dinate ratio indicating a more brittle behavior. Metals, the red, green and magenta

groupings in Figure 20, have higher yield and tensile strengths.

Figure 20. Yield strength vs. tensile strength for different materials [data courtesy of Granta Design Ltd].

Two temperature-related parameters are the thermal expansion coeffi cient and

the maximum service temperature. The thermal expansion coeffi cient describes how

much a material expands when being exposed to heat, and the maximum service

temperature is the highest temperature where the material keeps a majority of its

strength [97] and can be reasonably used for a prolonged period of time. In Figure

12, it becomes evident that metals are consistent in terms of thermal expansion, but

that the maximum service temperature diff ers greatly (this is shown by the fl at, wide

grouping of metals near the X-axis in fi gure 21). Polymers and polymer composites,

on the other hand, show no signifi cant conformity in either thermal expansion rate or

maximum service temperature (this is shown by the much larger, quadrangle shape

44

of the polymer grouping in Figure 21), even though the thermal expansion cofef-

fi cient is in general lower for metals, and the maximum service temperature is in

general higher for metals.

Figure 21. Thermal expansion coeffi cient vs. maximum service temperature for different materials [data courtesy of Granta Design Ltd].

A combination of strength and maximum service temperature is sometimes need-

ed to manage combined requirements. These requirements can occur when compo-

nents are stressed during increased temperatures.

Figure 22 shows the generalized material properties for all material classes, and it

can be seen that material properties can diff er greatly between materials and mate-

rial classes.

45

Diving deeper into material properties, looking at materials and alloys in particu-

lar (fi gure 23) there are vast diff erences even within this material class. Lead has a

signifi cantly lower maximum service temperature than nickel, with steel (both mild

and stainless) somewhere in between.

Figure 22. Strength over maximum service temperature. Both composites and polymers are shown to have lower possible maximum service temperature than metals and alloys. [96]

Figure 23. Strength over maximum service temperature for metals and alloys. The difference between differ-ent metals and alloys becomes evident. [96]

46

In the same fi gure 23, the diff erence in strength between diff erent metals is also

shown. As in the case of maximum service temperature, lead is at the lower end of

the scale and nickel is at the higher end. Aluminum and steel alloys have properties

in the mid to high range, with high-alloy steels being some of the absolutely strongest

metals.

Figure 24. Strength over maximum service temperature for polymers. [96]

Polymers have a much tighter span of maximum service temperature, and if foams

and rubbers are excluded, this is applicable for strength as well (see fi gure 24). The

maximum values are lower than for metals and alloys, but the diff erences within the

material class are much smaller. The diff erences in properties are as small among

polymer composites, but the composite materials are typically much stronger. The

maximum service temperature range, on the other hand, is very similar to the pure

polymers, as can be seen in fi gure 25.

47

Figure 25. Strength over maximum service temperature for polymer composites. CFRP and GFRP are shown to have similar maximum service temperature, but a difference in strength. [96]

48

3.2 STATE OF THE INDUSTRY

When discussing the introduction of new materials into the automotive industry,

some overview of industry trends is needed. While the automotive industry has used

lightweight materials as niche or low-volume solutions for a long time, some stake-

holders in the industry foresee an increased use of hybrid or multi-material designs

[98] in the near future. As an example, diff erent polymer-based materials have been

used for hang-on parts such as hatches, hoods and fenders for a long time [99], but

in the form of fi ber-reinforced polymer composites they have also recently been used

in roofs [100] and in more intricate multi-material designs of hang-on parts [101],

as can be seen in fi gure 26.

Figure 26. The tailgate of the 2017 Toyota Prius, exhibiting an integrated design with CFRP and aluminum [picture courtesy of Toyota].

49

The fi rst documented use of FRP body panels in a car dates back to 1941, when

Ford presented the “Soybean Car” [102]. Looking at mass-produced vehicles, two

pioneering cars were the Chevrolet Corvette, launched in 1953, and the Lotus Elite

(Type 14), launched in 1957 (as seen in fi gure 27). The Corvette used an FRP body

on a traditional steel frame [103], while the Elite utilized an FRP monocoque design

[104]. Approximately 35 years ago, in 1981, the fi rst car with a CFRP main load-bear-

ing structure was introduced with the McLaren MP4/1 Formula 1 car [105], and

since then the material has been used for load-bearing structures in diff erent types of

sports cars in parallel with the introduction of fi ber composites in hang-on parts in

mass production vehicles.

Figure 27. Lotus Elite (Type 14) [picture courtesy of ClassicarGarage / Marc Vorgers].

50

Currently, a number of diff erent vehicles are being released with diff erent types of

multi-material solutions even further integrated into the bodies. The trend seems to

be to introduce lightweight materials in more integrated designs, in low-cost products

and in transitioning from low production volumes to mass production (production

numbers of at least 10,000 units yearly). A few examples of such cars are presented

on the following pages, based on company-provided data and industry literature.

Audi TT Coupe

The 2014 model year and up Audi TT (see fi gure 28) has a body using stamped alu-

minum sheets for all outer body panels, while the fl oor and fi rewall are made out of

regular steel panels. Cast aluminum and aluminum profi les are utilized in safety-crit-

ical areas such as A-pillars, door sills and crash beams.

Figure 28. The Audi TT body, using aluminum and regular steel for a majority of the body structure [Picture courtesy of Audi AG].

51

Audi Q7

The 2015 model year and up Audi Q7 has an aluminum-intensive body, with both

cast and stamped aluminum integrated with high-strength steel in, for example,

B-pillars, as can be seen in fi gure 29.

Figure 29. The Audi Q7 body, using stamped aluminum sheets for most of its structural strength, combined with high-strength steel in safety-critical areas [Picture courtesy of Audi AG].

BMW 7 Series

The body of the 2015 model year and up BMW 7 Series [106] uses a combination of

steel, aluminum and carbon fi ber-reinforced plastics to reduce the mass of the BIW;

several components (for example door sills, B pillars and roof beams) have been re-

inforced or replaced with CFRP panels instead of the traditional steel panels. These

CFRP panels can be seen as the darker segment on the body in fi gure 30 & 31.

52

Figure 30. BMW 7 series carbon core structure [photo courtesy of BMW AG]. The CFRP B-pillar is prominent in the center of the picture, as well as the CFRP reinforcement panel behind the rear door opening.

Figure 31. BMW 7 series carbon core structure, the CFRP panels can be seen in black and aluminum panels in light silver grey [photo courtesy of BMW AG].

53

BMW i3

The BMW i3 (fi rst sold as a 2014 model year) is a small electric vehicle built up

in two major modules, the drive module (where the powertrain and suspension are

mounted) and the life module (where passengers are protected). The drive module

is made as an aluminum skateboard, while the life module is a CFRP monocoque as

seen in fi gure 32.

Figure 32. The BMW i3 showing both the drive module and the life module [picture courtesy of BMW AG].

Cadillac CT6

The Cadillac CT6 (2016 model year and up) is built on what General Motors calls the

“Fusion Frame”, a platform concept where a steel center section (marked in orange

in fi gure 33) is clad with aluminum panels for everything that can be seen to aff ect

the aesthetics of the vehicle (marked in grey in Figure 33). Aluminum is also used for

structural components such as door sills and impact bars.

54

Chevrolet Corvette

Ever since its introduction in 1953 (see fi gure 34), the Chevrolet Corvette has utilized

a fi ber-reinforced composite body. Initially, the body was made in glass fi ber-rein-

forced plastics, and the car was one of the fi rst mass-produced cars to have a com-

posite material body.

Figure 33. Cadillac CT6 “Fusion Frame” architecture [107]. The orange panels are of steel, while the grey panels and beams are made in aluminum [Picture courtesy of General Motors].

Figure 34. The 1953 Chevrolet Corvette, one of the fi rst mass-produced cars to have a body made from composite materials [Picture courtesy of General Motors].

55

In the beginning, the body was placed on a steel frame, but later versions have

transitioned into a unibody design. The lastest iteration, introduced as a 2014 model

year, uses as an aluminum center section (shown in fi gure 35) clad with composite

body panels.

Figure 35. The 2014 Chevrolet Corvette Stingray aluminum center section [Picture courtesy of General Motors].

Mercedes C-Class

For the model year 2016 C-Class (fi gure 36 and 37), Mercedes developed a body with

all hang-ons (doors, hood, fenders, trunk lid) in aluminum, along with an aluminum

roof and cast aluminum suspension mountings (marked in green in Figure 37). The

BIW also has a signifi cant amount of high-strength steel (hot-formed as well as reg-

ular), where torsional stiff ness is valuable.

56

Figure 36. The new Mercedes C-Class has a signifi cant amount of aluminum in its body panels, along with cast aluminum suspension mountings [picture courtesy of Daimler-Benz AG].

Figure 37. The new Mercedes C-Class has a signifi cant amount of aluminum in its body panels, along with cast aluminum suspension mountings and high-strength steel grades [picture courtesy of Daimler-Benz

AG].

57

Mercedes GLC

The Mercedes GLC (2016 model year and up) uses a combination of hot-formed

steel, ultra-high-strength steel alloys and aluminum alloys in addition to the clas-

sical mild to high-strength steel alloys in the body, as can be seen in fi gure 38. The

aluminum alloys are used in panels up front or high in the vehicle, possibly to move

the center of gravity backwards and downwards in the vehicle. Some additional alu-

minum parts are used in the front and rear suspension mountings.

Figure 38. The Mercedes GLC showing the use of hot-formed steel, ultra-high-strength steel and aluminum in the body [pictures courtesy of Daimler-Benz AG].

58

Volvo XC90

Volvo’s XC90 (2015 model year and up) utilizes high-strength steel and aluminum

components (both cast and stamped components) in order to reduce mass. Through-

out the BIW structure, multimaterial joints are present, as can be seen in fi gure 39.

Cast aluminum is used in the strut towers, while the hood and front quarters are

made from stamped aluminum plate.

Figure 39. The Volvo XC90’s various high-strength steel grade and aluminum hybrid structure [picture cour-tesy of Volvo Cars AB].

Summary of industry state

From the presented industry examples, some trends can be seen:

•Almost all passenger vehicle conglomerates have a product line or platform with light-

weight materials integrated

•Many companies try to incorporate lightweight materials in the upper parts of the body

•Multi-material designs are common

•Regular steel is exchanged for both high-strength steel and aluminum, depending on

the application

59

3.3 A GENERAL MODEL FOR PASSENGER VEHICLE PRODUCTION

In order to explain specifi c issues or challenges that occur when introducing new

materials in an existing production system, a generalized model of passenger vehicle

production is hereby presented (seen in fi gure 40). This generalized model is built

upon on-site observations and unstructured interviews with employees at four facto-

ries producing vehicles for four diff erent companies in three continents during 2014-

2016, as well as video observations and literature fi ndings from other passenger-ve-

hicle producing companies.

The model explains the production fl ow from the stamping of body panels to the

fi nished product ready for quality assurance testing before delivery. The fl ow follows

the path of the main structure of the unibody structure, since this is commonly ac-

cepted as the identity carrier of the vehicle (as an example, the VIN tag is seen as the

ID tag of the vehicle; this is usually joined to the unibody structure in some way).ID tag of the vehicle; this is usually joined to the unibody structure in some way).

Figure 40. Generalized vehicle production model.

The process begins in the body stamping shop, where panels are formed and cut

to size to be able to be used in later work stations (see fi gure 41). Some components

are joined, such as the inner and outer panels of the hood and doors. The process is

mainly parallelized and made in batches to reduce the downtime needed for switch-

ing tool sets.

60

Figure 41. Picture of a body stamping shop [Picture courtesy of Volvo Cars AB].

The panels are transported to the body shop where they are distributed to specif-

ic work stations, and panels are welded together to start building the unibody (see

fi gure 42). Enclosures (hood, trunk, doors) and their hinges are mounted, and some

panels are screwed onto the unibody (usually front fenders). Here, the fl ow could be

specialized, batched or mixed-model.

Figure 42. Automated unibody welding with welding robots in a body shop [picture courtesy of Volvo Cars AB].

61

The unibody is moved into the paint shop, where some panel seams are covered

with fi lling material, before the body is coated with anti-corrosion protectives and

painted with primer and topcoat (see fi gure 43). The unibody is then run through a

curing oven, where the body is heated up to approximately 200 degrees Celsius for

some time to cure the paint before it is polished and possible quality issues are recti-

fi ed. Here, the fl ow could be specialized, batched or mixed-model.

Figure 43. Automated topcoat painting in the paint shop [picture courtesy of Volvo Cars AB].

62

The painted unibody is moved into the assembly shop (see fi gure 44), where sub-

systems are assembled to the unibody to fi nish the car. The subsystems assembled

can be classifi ed into several categories: powertrain, chassis and suspension, elec-

tronics, interior and glasswork. Sometimes, multiple subsystems (such as powertrain

and chassis and suspension) are assembled or fi xed together to make a larger subsys-

tem for a “marriage” with the unibody. Here, the fl ow could be specialized, batched

and mixed-model.

Figure 44. Final assembly of a modern passenger vehicle [picture courtesy of Volvo Cars AB].

63

PART 4 - CONTRIBUTION

64

PART 4 - CONTRIBUTION

In the contribution, the work performed by the author is presented. First, a tech-

nical problem analysis is performed. After this, the four main cases are presented

and an analysis of all the work performed is put forward in order to form the fi nal

conclusions. This part of the thesis should be read as the author’s contribution to the

academic knowledge base. Parts of the cases presented have been presented earlier

in papers B and C.

Picture on previous page courtesy of Philip Ekströmer

65

4.1 INDUSTRIAL PROBLEM ANALYSIS

The research in this thesis aims to not only contribute to the academic knowledge

base, but also contribute with knowledge that can be implemented in an eff ort to

solve societal challenges. The review and defi nition of the industrial problem is done

in two ways: one going backwards from the societal challenge of energy effi ciency

and making a technical problem analysis, and one analyzing the current situation in

research on product development processes in an organizational problem analysis.

Technical problem analysis

If the challenge is to increase energy effi ciency in road vehicle transportation, this

challenge can be further investigated by defi ning energy consumption and energy

effi ciency in road vehicles.

Looking at a car driving along a road in a straight line (as in fi gure 45), the fuel

consumption of this vehicle can be calculated using Equation 2. The energy con-

sumption of a battery electric-powered vehicle would be calculated in a similar man-

ner, using some smaller modifi cations.

66

Looking further into Ft, the tractive force, an explanation can be seen below in

Equation 3 [109, 110]. Mass can be seen to contribute to the rolling resistance, accel-

eration resistance and climbing resistance [109, 110]. Since few roads are perfectly

horizontal, and much driving consists of acceleration or deceleration, all these resist-

ances aff ect much of the driving in our society. So if vehicle mass can be reduced, each

vehicle will either transport more payload for equal fuel consumption or have a lower

fuel consumption per travelled distance.

Figure 45. A passenger vehicle driving along a road.

67

Lightweighting

Lightweighting is the activity of reducing mass in the end product. Lightweighting,

or mass reduction, is recognized by Ulrich and Ellison [111] as one of the main driv-

ers for developing bespoke solutions as opposed to using existing components and

technology. The defi nition of material density is described as mass divided by vol-

ume, as seen in Equation 4:

However, in product development and engineering design work, density is seldom

the resulting quality. This is especially true in land-based products, where buoyancy

is not a defi ning quality of the product. Instead, requirements are usually based on

the mass of the component. That suggests a restructure of Equation 5 to say:

Mass reduction (from generation n to generation n+1 of components or products)

could be described as:

68

Equations 5 and 6 show that a reduced mass in a component can be achieved ei-

ther by a reduction in volume (vn+1 < vn) or a reduction in density (ρn+1 < ρn), or a

combination of both. In the extreme case, the entire mass reduction is achieved by

switching to a material with lower density or by removing material from an existing

design. This also means that mass reduction could be described as an optimization

problem, where we are searching for

for a set of boundary conditions aff ecting the possible values of V and ρ. These bound-

ary conditions are constructed by the customer requirements, product development

processes, supply chains, production system restrictions, product cost and after-sales

considerations. These boundary conditions are interdependent and non-static, which

means that the at fi rst the simple optimization problem from Equation 7 becomes

a complex and complicated optimization problem, and thus it makes engineering

sense to investigate non-mathematical ways of solving the problem. Some of these

non-mathematical ways of solving the optimization problem could also increase

knowledge about the boundary conditions for the mathematical solution.

Organizational problem analysis

Product development processes can be classifi ed into two categories, linear and

non-linear processes.

Linear product development processes have issues with ambiguous data input into

the process, and ambiguous data can halt a product development project or kill off

a project that might have been successful had it not been for decisions made on this

ambiguous data.

69

Non-linear product development processes need signifi cant amounts of data and

projects that can be segmented into small increments, either by having existing data

(as in systems engineering, especially suited for the development of complex systems

where precedents exist) or by being extremely modular (as in agile methodologies,

especially suited for software development where each function or subscript can be

tested individually).

Concluding problem analysis

A way of increasing energy effi ciency in a passenger vehicle is to reduce the mass.

This could be described as the optimization problem

with boundary conditions on V and ρ describing among other things the product

development process, product requirements and production system restrictions. Due

to the complexity of these boundary conditions, a non-mathematical approach to

solving the problem should be investigated. This could be done by investigating the

process in which the products are created.

Both existing classes of product development processes have challenges when

working with the type of problem that the introduction of new materials is (am-

biguous data input, lack of relevant data, complex systems of systems and lack of

modularity). This means that there is not a clear-cut solution to what the product

development process should look like.

70

4.2 CASE STUDIES

Within the research work presented in this thesis, four case studies have been exec-

tured. These four case studies will be presented on the following pages.

Case 1 – Carbon fi ber roof panel

The fi rst case, Case 1, was set